INDEX

Explanations

phrases or words related to opinions or categorizations, particularly when something is being perceived or labeled in a certain way

oai_token-act-pair · gpt-3.5-turbo

terms that indicate judgment or classification of people, ideas, or phenomena

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ajo

-0.75

load

-0.57

 planes

-0.55

 occupancy

-0.55

 purse

-0.54

Zone

-0.54

holder

-0.54

 hearts

-0.53

ipeg

-0.53

 Lies

-0.53

POSITIVE LOGITS

escription

0.82

 favorably

0.78

Ô

0.74

uted

0.72

ikers

0.72

Ú

0.71

STDOUT

0.70

ï¸

0.69

maxwell

0.68

entious

0.65

Activations Density 0.138%

phrases or words related to opinions or categorizations, particularly when something is being perceived or labeled in a certain way

terms that indicate judgment or classification of people, ideas, or phenomena

No Comments

No Known Activations